Comparison of the Validity and Generalizability of Machine Learning Algorithms for the Prediction of Energy Expenditure: Validation Study.
Comparison of the Validity and Generalizability of Machine Learning Algorithms for the Prediction of Energy Expenditure: Validation Study.
复制标题
DOI：
10.2196/23938
复制
发表时间：
2021-08-04
影响因子：
5
通讯作者：
Stubbs RJ
中科院分区：
文献类型：
--
作者：
O'Driscoll R;Turicchi J;Hopkins M;Duarte C;Horgan GW;Finlayson G;Stubbs RJ
一系列医疗和健康研究领域需要准确的解决方案来大规模估计身体活动和能量消耗。机器学习技术在研究级加速度计中显示出前景，一些证据表明这些技术可以应用于更具可扩展性的商业设备。本研究旨在使用包含不同活动的两个实验室数据集来测试预测多种可穿戴设备（即 Fitbit Charge 2、ActiGraph GT3-x、SenseWear Armband Mini 和 Polar H7）能量消耗的算法的有效性和样本外泛化性。本研究合并了两项实验室研究（研究 1：n=59，年龄 44.4 岁，体重 75.7 kg；研究 2：n=30，年龄=31.9 岁，体重=70.6 kg），其中成年参与者执行了基于实验室的连续活动方案，包括休息、家庭、走动和非走动任务。在这两项研究中，都使用间接量热法从可穿戴设备收集加速度计和生理数据以及能量消耗数据。三种回归算法用于预测代谢当量（MET；即随机森林、梯度提升和神经网络），五种分类算法（即 k 最近邻、支持向量机、随机森林、梯度提升和神经网络）用于将身体活动强度分类为久坐、轻度或中度到剧烈。使用留一受试者交叉验证和样本外验证来评估算法。应用于 SenseWear 和 Polar H7 数据的梯度提升的均方根误差 (RMSE) 最低 (0.91 MET)，并且在分类任务中，应用于 SenseWear 和 Polar H7 的梯度提升最准确 (85.5%)。 Fitbit 模型的 RMSE 为 1.36 MET，分类准确度为 78.2%。 SenseWear 神经网络在回归任务中实现了 1.22 MET 的 RMSE 值，而 SenseWear 梯度提升和随机森林在分类任务中实现了 80% 的准确率，样本外验证中的错误趋于增加。在组合数据集上训练的算法表现出很高的预测准确性，对于大多数但并非所有可穿戴设备，随机森林和梯度提升都有优越的性能。研究间验证的预测较差，这给测试算法的普遍性带来了不确定性。
Accurate solutions for the estimation of physical activity and energy expenditure at scale are needed for a range of medical and health research fields. Machine learning techniques show promise in research-grade accelerometers, and some evidence indicates that these techniques can be applied to more scalable commercial devices. This study aims to test the validity and out-of-sample generalizability of algorithms for the prediction of energy expenditure in several wearables (ie, Fitbit Charge 2, ActiGraph GT3-x, SenseWear Armband Mini, and Polar H7) using two laboratory data sets comprising different activities. Two laboratory studies (study 1: n=59, age 44.4 years, weight 75.7 kg; study 2: n=30, age=31.9 years, weight=70.6 kg), in which adult participants performed a sequential lab-based activity protocol consisting of resting, household, ambulatory, and nonambulatory tasks, were combined in this study. In both studies, accelerometer and physiological data were collected from the wearables alongside energy expenditure using indirect calorimetry. Three regression algorithms were used to predict metabolic equivalents (METs; ie, random forest, gradient boosting, and neural networks), and five classification algorithms (ie, k-nearest neighbor, support vector machine, random forest, gradient boosting, and neural networks) were used for physical activity intensity classification as sedentary, light, or moderate to vigorous. Algorithms were evaluated using leave-one-subject-out cross-validations and out-of-sample validations. The root mean square error (RMSE) was lowest for gradient boosting applied to SenseWear and Polar H7 data (0.91 METs), and in the classification task, gradient boost applied to SenseWear and Polar H7 was the most accurate (85.5%). Fitbit models achieved an RMSE of 1.36 METs and 78.2% accuracy for classification. Errors tended to increase in out-of-sample validations with the SenseWear neural network achieving RMSE values of 1.22 METs in the regression tasks and the SenseWear gradient boost and random forest achieving an accuracy of 80% in classification tasks. Algorithms trained on combined data sets demonstrated high predictive accuracy, with a tendency for superior performance of random forests and gradient boosting for most but not all wearable devices. Predictions were poorer in the between-study validations, which creates uncertainty regarding the generalizability of the tested algorithms.
登录
查看更多内容
影响因子：
4.7
作者：
Brage, S;Brage, N;Wareham, NJ
通讯作者：
Wareham, NJ
影响因子：
3.3
作者：
Brage, S;Brage, N;Wareham, NJ
通讯作者：
Wareham, NJ
影响因子：
3.4
作者：
Kim, Dohyun;Lee, Jongshill;Kim, In Young
通讯作者：
Kim, In Young
DOI：
10.1186/s12966-014-0119-7
发表时间：
2014-09-28
影响因子：
8.7
作者：
Calabro, M. Andres;Lee, Jung-Min;Welk, Gregory J.
通讯作者：
Welk, Gregory J.
影响因子：
3.6
作者：
CEESAY, SM;PRENTICE, AM;SPURR, GB
通讯作者：
SPURR, GB